Learning to communicate using natural language is a uniquely human capacity that underlies the exchange of information among people. There is as yet no consensus concerning how susceptible this process is to environmental and biological variation. The proposed Program Project addresses this issue, with a focus on how language is used for higher order thinking, a key cognitive underpinning of academic and 21st century career success. To examine the impact of environmental variation on language used for higher order thinking (Project I), 60 children, who were selected to represent the demographic range of the Chicago area and had been observed between 14 months and 10 years during the previous grant periods, will be followed until 15 years as they enter early adolescence. To explore continuity over development, higher order thinking will be coded in videotapes of spontaneous parent-child conversations early in development, and in tasks designed to elicit higher order thinking that parents conduct with their children at ages 10, 12, and 14 years. Standardized assessments of higher order thinking will also be collected. Using these data, models of cumulative parent input to higher order thinking from 14 months to 10 years will be constructed and used to predict children's subsequent skills in higher order thinking. To explore relations between language and higher order thinking in the face of biological variation (Project II), 40 children with unilateral brain injury, who had been observed from 14 months and will now be followed from 11 to 15 years, will be studied to determine how environmental input and biological aspects of their lesions combine to predict their use of language for higher order thinking, and whether input plays a similar role in this group as it does for children who have not suffered brain injury. In addition, the brain bases underlying language used for higher order thinking will be assessed in children from Projects I and II using fMRI techniques (Project III). Three cores will provide broad support to the projects: Administrative Core A, Data Collection and Transcription Core B, and Statistical Core C. The proposed work builds on and extends 10 years of longitudinal data in a diverse sample of typically developing children and children with brain injury, and thus offers a unique opportunity not only to explore the impact of early parental input on the development on higher order thinking, but also to examine how that input interacts with early brain injury.